The initial focus of this grant was paraneoplastic immunity to Ma proteins. We initially identified this disorder in men with testicular tumors and brainstem and limbic encephalitis (LE: memory deficits, seizures, sleep disorders) who developed antibodies to neuronal proteins expressed by the tumors. These antibodies served to isolate the target antigens (Mal-3, a family of brain-testis proteins), assisted in characterizing the disorder (Maencephalitis), and led to the development of diagnostic tests currently used worldwide. As this work proceeded we identified patients with treatment responsive LE who were initially thought to have Ma antibodies on clinical grounds but when tested, had other antibodies. An important discovery has been the identification of a new category of autoimmune encephalitis and the target autoantigens, leading to successful treatment strategies for this and similar disorders. The affected patients are young women who develop prominent psychiatric symptoms, seizures, and central hypoventilation often requiring prolonged ventilatory support. All of these patients had antibodies to cell surface antigens (highly enriched in hippocampus), which we identified as the NR2B (and at a lesser degree NR2A) subunits of the NMDA receptor. Although this encephalitis is more lifethreatening than other paraneoplastic disorders, patients are more responsive to treatment (with frequent full recovery) if properly diagnosed. Extensive preliminary data support the concept that the neurologic and psychiatric symptoms are mediated by the NMDA receptor (NMDAR) antibodies. In addition, we identified other treatment responsive phenotypes (some without a cancer association) that occur with antibodies to other cell surface neuronal antigens (pending characterization). Based on these data we postulate the occurrence of two large and heterogeneous groups of autoimmune LE; one group (prototype: Ma-like proteins) includes disorders related to intracellular antigens; the other group (prototype: NMDA receptor) includes disorders related to cell membrane antigens. While disorders of the first group associate with cancer, brain cytotoxic T-cell infiltrates, and variable response to treatment, the second group of disorders associate less frequently with cancer, appear to be mediated by antibodies, and usually respond to IgG-depleting strategies. The three specific aims of this competing renewal application are to 1) fully define the clinical-immunological phenotypes of LE and repertoire of NR2 (NMDAR) antigens, 2) map the epitope regions of NR2 subunits of the NMDAR and isolate other hippocampal antigens, and 3) determine whether antibodies to NR2 subunits of the NMDAR cause LE in a mouse model.